Date of Award

Spring 5-15-2020

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Neurosciences)

Degree Name

Doctor of Philosophy (PhD)

Degree Type



Alzheimer’s disease (AD) is a neurodegenerative disorder associated with irreversible damage to the brain, which manifests in cognitive dysfunction, memory loss, and eventual death. The pathological hallmarks of AD are amyloid plaques, which are cerebral aggregates consisting of fibrils of the amyloid β-protein (Aβ), and filamentous lesions of the microtubule-associated protein tau known as neurofibrillary tangles. In the early 1990s, the apolipoprotein E (apoE) was found to co-localize with amyloid plaques. The ε4 allele of the APOE gene was sequentially identified as the strongest genetic risk factor for AD, increasing the risk by 4 – 12-fold, whereas the ε2 allele is protective relative to the prevalent ε3 allele. Since then, multiple lines of evidence suggest that the major mechanism by which apoE influences AD pathology is via its effects on Aβ metabolism, particularly aggregation and clearance. An ongoing debate in the field is whether the ε4 allele impose a loss of protective function or a gain of toxic function. In support of the latter hypothesis, our colleagues previously demonstrated that APP/PS1-21 mice with only one copy of human apoE3 or apoE4 have significantly less amyloid plaque deposition and microglial activation compared to their homozygous littermates. However, the effect of apoE reduction during the post-natal period or adulthood is unknown. To address this gap in knowledge, we utilized an apoE antisense oligonucleotide (ASO) to reduce apoE expression in the adult APP/PS1-21 mice homozygous for the human ε4 allele of APOE. Despite achieving reduction of apoE expression by more than 50% starting at the onset of amyloid deposition, no reduction of Aβ pathology was detected when mice were assessed at 4 months of age. Though there was not an overall reduction in amyloid deposition, there was a clear effect of reducing apoE4 on Aβ plaque morphology. Interestingly, ASO treatment starting after birth led to a strong and significant decrease in Aβ pathology when mice were assessed at 4 months of age. These results suggest that apoE levels can strongly affect the initiation of Aβ pathology in vivo but that once Aβ plaque pathology is present, reducing apoE does not have a strong effect on further amyloid deposition. This previously unknown age-dependent effect of apoE in the early stages of Aβ plaque formation suggest the important implication that decreasing brain apoE levels would be useful for primary prevention of amyloid deposition but not for decreasing or removing amyloid plaques once they have begun depositing. Strikingly, we observed a marked decrease in neuritic dystrophy around the plaques in APP/PS1-21/ε4 mice treated with ASO under either treatment paradigm, independent of plaque size or plaque load. This suggests a general role of apoE4 in modulating the brain’s response to neurotoxic insults (such as Aβ plaques), independent of its effects on Aβ metabolism. The The aggregation of Aβ into higher-order species follow nucleation-dependent kinetics in vitro. Our work suggested apoE affects the earliest stages of plaque formation (the nucleation phase), but it remains unclear whether apoE isoforms exert differential effects. We utilize an established in vivo seeding protocol to investigate the possibility that apoE can influence the formation and/or potency of the Aβ seeds in an isoform-dependent manner. We inoculated PBS-soluble brain extracts (containing Aβ seeds) isolated from aged APP/PS1-21 donor brains expressing different human APOE alleles (ε2/ε2, ε3/ε3, ε4/ε4) in the hippocampus of an APP-expressing host. Following a defined incubation period, we analyzed the seeding patterns and found that brain extracts from APP/PS1 donors with different APOE backgrounds induce Aβ seeding patterns that are distinct from each other. Specifically, seeded Aβ species from ε2 donors have a diffuse pattern with minimal fibrillar content, while those from ε4 donors appear more punctate-like and are mostly fibrillar. Brain extracts from ε3 mice produced plaques with an intermediate phenotype. These results suggest that human APOE isoforms may differentially affect the properties of Aβ seeds, thus creating different strains of Aβ with distinct structural features and seeding capabilities. This isoform-dependent effect of apoE on Aβ may contribute to the overall AD risk associated with the different APOE isoforms. Further studies are needed to investigate the consequences of this isoform-specific difference in plaque morphology. As apoE is produced both inside and outside of the central nervous system (astrocytes and microglia in the brain, and hepatocytes in the periphery), the specific contributions of different apoE pools to AD pathogenesis remain unknown. To address some aspects of this question, we generated new lines of APOE knock-in (APOE-KI) mice (ε2/ε2, ε3/ε3, and ε4/ε4) where the exons in the coding region of APOE are flanked by loxP sites, allowing for cell type-specific manipulation of gene expression. We assessed these mice both alone as well as after crossing them with mice with and without amyloid deposition in the brain as well as after removing apoE expression from hepatocytes using biochemical and histological methods. Consistent with prior studies, our analyses demonstrated apoE protein was present predominantly in astrocytes in the brain under basal conditions and was also detected in reactive microglia surrounding amyloid plaques. Primary cultured astrocytes and microglia from the APOE KI mice secreted apoE in lipoprotein particles of distinct size distribution upon native gel analysis with microglia particles being substantially smaller than the HDL-like particles secreted by astrocytes. Crossing of APP/PS1 transgenic mice to the different APOE-KI mice recapitulated the previously described isoform-specific effect (ε4 > ε3) on amyloid plaque and Aβ accumulation. Deletion of APOE in hepatocytes did not alter brain apoE levels but did lead to a marked decrease in plasma apoE levels and changes in plasma lipid profile. Despite these changes in peripheral apoE and on plasma lipids, cerebral accumulation of amyloid plaques in APP/PS1 mice was not affected. Altogether, our new APOE knock-in strains offer a novel and dynamic tool to study the role of APOE in AD pathogenesis in a spatially and temporally controlled manner.


English (en)

Chair and Committee

David M. Holtzman

Committee Members

Marco Colonna, Joseph M. Dougherty, Celeste M. Karch, Timothy M. Miller,